SUMMARY This research proposal is in response to AHRQ (NOT-HS-16-009) and examines the quality and safety impacts of real-time patient-provider video and telephone visits, interactive health information technology (IT) tools with potential to engage patients through convenient health care access. Telemedicine can offer patients the choice to access a provider visit without the obstacles of securing transportation, arranging time-off from work or care-giving, or spending time in a waiting-room. Even as American patients are increasingly purchasing internet-based telemedicine visits from third party services without in-person facilities, gaps remain in the scientific evidence about patient adoption of telemedicine for primary care encounters (compared with in-person visits). This study will examine a 2016 health IT implementation offering patient - initiated primary care telemedicine encounters through the patient portal of an integrated healthcare delivery system. All patients scheduling an appointment with a primary care provider through the portal (website and mobile application) choose directly between a traditional in -person visit and a telemedicine visit, either by video or telephone. Over a five year study period (2016 -2020), we will examine a large sample of patient-scheduled telemedicine appointments, including over 50,000 video visits and 500,000 telephone visits, compared with millions of patient-scheduled in-person visits. Using patient surveys (N=1,500) and stakeholder key informant interviews, we will examine the experiences of telemedicine users and decision-makers, including technology usability, convenience, and patient-reported outcomes. In the study setting, this novel patient telemedicine access is fully integrate d with patients? own existing health care providers and comprehensive electronic health record (EHR). We will examine patient clinical concerns and access measures associated with the choice of a telemedicine encounter compared with in-person encounters (Aim 1), and the quality and safety of the telemedicine encounters through guideline-recommended health care processes (prescribing and ordering), follow-up visits, and events ( emergency department visits) (Aim 2). We will test the hypotheses that care processes and short-term event rates for patient-initiated telemedicine encounters will not be worse than for patient-initiated in-person visits (non-inferiority hypothesis), while accounting for patient engagement, recent changes in health status, case-mix, cost-sharing for in-person visits, internet access, and patient demographic and socio-economic characteristics. By examining an early adopter of the technologies, and a large and diverse patient population, this project has the potential to provide timely evidence to inform emerging telehealth policies, technology adoption decisions, and real world use by patients and clinicians.